Home / How to scan documents to PDF

Guide · 2026

How to scan documents to PDF on an iPhone

Everything that separates a usable scan from a photograph of a piece of paper: lighting, geometry, filters, OCR and the filing habit that makes the whole thing worth doing.

Updated 1 September 2026 · About 6 minutes to read

Almost everyone has taken a photo of a document and emailed it. It works, in the sense that the recipient can usually read it. It also produces a file that is crooked, four times larger than it needs to be, grey down one side, impossible to search, and awkward to print. The gap between that and a proper scan is not equipment — the camera in your pocket is more than capable — it is a handful of small decisions made before and after the shutter.

This guide covers those decisions. It is written around PDF Scanner, but the physical technique applies to any scanning app.

Why a phone scan beats a phone photo

Three things happen in a scanning app that do not happen in the camera app:

  • Edge detection finds the boundary of the page and crops everything outside it, so you get the document rather than the document plus your kitchen table.
  • Perspective correction takes the trapezoid that a tilted camera produces and warps it back into a rectangle. This is why a scan looks flat even when you were leaning over the page.
  • Tone mapping pushes the paper towards pure white and the ink towards pure black, removing the grey cast that ambient light leaves behind.

Together these turn an image into a document. The rest of this guide is about giving those three processes the best possible input.

Step by step

  1. Set the page down on a contrasting surface. Put white paper on a dark table or a dark document on a light one. Edge detection works by finding the boundary between the page and what is behind it, so contrast is the single biggest factor in whether it locks on.
  2. Get the light in front of you, not behind. Stand so the window or lamp is facing the page rather than behind your shoulders. That keeps your own shadow off the paper, which is the most common cause of a grey band across a scan.
  3. Hold the phone parallel to the page. Keep the camera flat above the document rather than tilted. Perspective correction can fix a moderate angle, but it works from a better starting point when the frame is roughly square to the paper.
  4. Let the frame lock, then capture. Wait for the detected border to settle before capturing. A capture taken mid-adjustment is where crooked crops come from.
  5. Add the remaining pages. Keep shooting to append pages to the same document. Multi-page capture in one pass is far quicker than merging separate files afterwards.
  6. Apply a filter. Black and white for text-only pages, greyscale for pages with photographs, contrast boost when the lighting was uneven. This is what makes a phone photo look like a scan.
  7. Run OCR. Text recognition turns the image into searchable, selectable text so the document can be found later by its contents rather than its filename.
  8. Name it, file it, back it up. Give it a name you would actually search for, drop it into a folder, add a tag, and let iCloud take a copy.

Lighting: the thing that matters most

If you fix one variable, fix this one. Every other problem — failed edge detection, poor OCR accuracy, grey backgrounds, muddy contrast — traces back to light more often than to anything else.

The ideal is soft, even, indirect daylight: a page on a table a metre or so from a window on an overcast day. What you want to avoid is directional light that creates a bright side and a dark side across the same page, and your own shadow, which is what happens when the light source is behind you and you lean over the document.

Phone flash is a last resort. It produces a hot spot in the middle of the page and falls off sharply at the corners, which tone mapping then exaggerates. A cheap desk lamp positioned to one side, angled across the page rather than straight down at it, gives a much better result than flash.

Quick test. Before capturing, look at the live preview rather than the paper. If you can see a shadow gradient across the page on screen, move — the scan will bake that gradient in.

Choosing the right filter

Filters are not cosmetic. They change the file size, the print quality and how well OCR performs.

Document typeFilterWhy
Printed text: contracts, forms, lettersBlack & WhiteMaximum contrast for OCR, smallest file size, cleanest print
Receipts, especially faded thermal paperContrast boostRecovers detail without crushing light print to nothing
Pages mixing text and photographsGreyscaleKeeps tonal range in the images while still cleaning the paper
Anything where colour carries meaning: highlighted text, ID cards, chartsColour / noneColour is information — do not throw it away to save space

Multi-page documents

Capture every page in a single session rather than scanning each one separately and merging later. It is faster, it keeps the pages in order automatically, and it avoids the situation where you end up with eight one-page PDFs and no idea which is page three.

If a page comes out badly, retake just that one rather than starting over. And if you did end up with separate files, the Merge tool will combine them — but treat that as a repair, not a workflow.

OCR: making the document findable

A scan without text recognition is a picture of words. You cannot search it, you cannot copy from it, and in six months you will be scrolling thumbnails looking for it.

Running OCR embeds a text layer behind the image. The page still looks identical, but the words are now data. That means you can find the invoice by searching for the amount printed on it, or the medical form by the clinic's name — even if the filename is Scan 14 Jul.

Accuracy depends on input quality, which is why OCR comes after filtering rather than before. Printed body text in a common font is recognised near-perfectly. Handwriting, script fonts, very small print, and text over busy backgrounds are all harder. There is more detail in how OCR works on iPhone.

Naming and filing

The scanning is the easy part. The reason people give up on digital paperwork is that they end up with 400 documents called Scan and no structure.

Two habits fix this permanently. First, name the file at the moment you create it, using words you would actually search for later — Rent contract 2026, not Document 12. Second, put it in a folder immediately rather than promising yourself you will tidy up on Sunday. A full filing scheme, including a tag strategy that survives having hundreds of documents, is covered in organising documents on iPhone.

Backing it up

A scanned document that exists only on the phone in your pocket is not archived, it is temporarily stored on a device that gets dropped, stolen and replaced. Turn on iCloud backup so the library exists in a second place, and export genuinely irreplaceable documents — deeds, certificates, contracts — to a computer or external drive as well. The rule is simple: anything you would be upset to lose should exist in two places you control.

Frequently asked questions

Is a phone scan good enough for official use?

For the overwhelming majority of purposes — submitting a form, sending an invoice, filing a receipt, emailing a signed contract — yes. A well-lit phone scan with edge correction and a black-and-white filter is visually equivalent to a flatbed scan at 200–300 DPI. The exceptions are contexts that specifically demand a certified or notarised copy, or archival reproduction of artwork, where a dedicated scanner is still the right tool.

What resolution does a phone scan produce?

A modern iPhone captures a 12-megapixel image. Spread across an A4 page that works out to roughly 300 DPI, which is the standard threshold for print-quality document reproduction. Sharpness matters more than raw megapixels: a steady, well-lit capture at arm's length beats a shaky close-up every time.

How do I scan a book or a bound document?

Press the page as flat as you can and capture one page at a time rather than a two-page spread. Curved pages near the spine cause text distortion that OCR struggles with. If the binding will not open flat, photograph each page at a slight angle towards the spine and let perspective correction do the rest.

Why is my PDF file so large?

Colour scans of photographs carry far more data than text. Applying the black-and-white filter to text-only pages typically cuts file size by 70–90 per cent, because the image collapses to two tones. If you need to email a large multi-page document, filter it first.

Can I scan a receipt that is crumpled or faded?

Flatten it under a book for a few minutes first — creases cast shadows that read as page edges. For thermal receipts that have faded, a contrast-boost filter recovers more than you would expect, though OCR accuracy on faded thermal print is always going to be lower than on laser print.

Does scanning work in a dark room?

Poorly. The camera compensates for low light by raising ISO, which introduces noise, and noise is what breaks both edge detection and OCR. If you cannot get to daylight, put the page directly under a desk lamp rather than relying on the phone flash, which produces a hot spot in the centre of the page.

Try it on the pile you already have. PDF Scanner is free on the App Store — edge detection, filters, OCR and folders included, no account required.